Scaling Data Science at Zespri International
Brett Hartman, Head of Data and Analytics, Zespri International.
Zespri is 100 percent owned by current and former kiwifruit growers and has a global team of 800 based in Mount Maunganui and throughout Asia, Europe and the Americas.
The company’s purpose is to help people, communities and the environment thrive through the goodness of kiwifruit, and it works with 2,800 growers in New Zealand and 1,500 growers offshore to provide consumers with fresh, healthy and great-tasting Zespri Green, Zespri RubyRed™ and Zespri SunGold Kiwifruit. As part of this, Zespri manages an integrated production and distribution system used to deliver the world’s best kiwifruit to consumers worldwide. This Zespri System is the culmination of many years of scientific, technical, and practical developments and an uncompromised commitment to continual improvement.
Zespri is dedicated to its customers, consistently providing excellent products and service. They recognise that a consistently high level of customer service can only be attained by an integrated Supply Chain that, from orchard to point of retail sale, is focused on understanding and meeting consumer’s needs.
Zespri asked Ignite to help them design a unique Data Science Operating Model. They wanted a complete review and refresh of the existing data science roadmap, detailed processes on how the data science practice will work, how it can scale with partners, and be socialised with and accepted by the Zespri Executive Team. Importantly, Zespri wanted to be sure that its unique IP is protected and existing solutions are re-used – avoiding repeat effort in creating solutions from scratch.
“Zespri gathers data throughout the product value chain, from the orchard through to the consumer. We are always thinking – ‘how we can use this data to make better decisions?’ There have been significant advances in areas such as predictive algorithms, and our team focuses on how we can optimise data science and advanced analytics to drive competitive advantage. We are constantly looking to do things more efficiently, make better operational decisions and understand our consumers at a far deeper level,” says Brett Hartman, Zespri’s Head of Data, and Analytics.
“Over time, different parts of the business had created their own initiatives and projects to address their immediate needs. We hadn’t taken an overall business approach. This meant that we had some siloes and were missing opportunities to optimise. What we are really trying to do with Ignite is to position data science at the exec level, so we have the right governance and structure, and we can hard wire re-use across the business.”
The primary purpose of the Zespri Data Science Operating Model is to reduce the risk of informally doing business by representing how Zespri can efficiently and effectively deliver the potential value of data through data science.
It describes the data science lifecycle, the people involved, the core technology, the delivery model, reporting on effort and return, and a roadmap of activities. Its focus is on maximising business value through top-line growth and reducing business risk.
One of the best things we developed with Ignite is a process to evaluate ideas across the business. Ignite brings a real people focus to process, they think about how we can embed good ideas into our everyday.
“Through this process, we have put structure and process around what we do in the Data Science team. It covers how we bring initiatives into our pipeline of activities, how we can prove the value, and then operationalise these initiatives within our businesses. We look at people, processes and data. We now have a considered and defined process to manage our ideas pipeline. We have developed the model, and now we will start to implement it,” says Zespri Data Scientist, Stuart Fish.
“Ignite was recommended to us. There are not that many companies out there in this space, other than the big consulting firms that tell you they can do anything but can’t always! We were looking for more of a partnership,” says Brett. “Ignite has delivered an operating model that we’re confident is going to deliver benefits not only in the short-term but in the years ahead.”
Embedding the end-to-end process into Zespri’s business from how they use their systems through experimentation, all the way to embedding them into everyday activities and monitoring the results. Whilst encouraging reuse through the core Zespri Analytics Platform powered by Azure Databricks.
“We are working with a smaller specialist partner who has a good understanding of what we are trying to achieve. We focused on a two-speed environment – experimenting quickly in a prototype/prove value way and then implementing a robust process to operationalise models. Really proving that you can apply data science to a problem, is more than just that initial experiment – it is about how you can embed that into the business and create and count value. The model looks at the whole business process approach and Ignite brought that focus.” – Stuart Fish, Data Scientist, Zespri International.
We believe the best results come when data enables people
Let's put data at the heart of improving your business and communities you serve.